INDEX

Explanations

specific mentioned actions or requirements within a sequence of actions

oai_token-act-pair · gpt-3.5-turbo

verbs and their forms within conditional or explanatory contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

erva

-0.77

 tracts

-0.68

 Nanto

-0.67

 grades

-0.67

 finance

-0.66

 finances

-0.65

 traject

-0.65

 Sons

-0.65

 Buchanan

-0.64

 Bundy

-0.62

POSITIVE LOGITS

 optional

1.27

 annoying

1.18

 buggy

1.17

 handy

1.12

 implemented

1.12

 useful

1.10

 disabled

1.09

 enabled

1.06

 nifty

1.05

 deprecated

1.05

Activations Density 0.262%

specific mentioned actions or requirements within a sequence of actions

verbs and their forms within conditional or explanatory contexts

No Comments

No Known Activations